The increasing number of companies shutting down or restricting access to formerly open APIs has alarming implications. This is concerning, not only in the case of extremist CEOs hiding the inner workings of their platforms or companies putting up substantial paywalls for access as the author suggests, but also as it hinders research and innovation that would have come from the previously open APIs.

The idea of "good AI" is a very hopeful thought, but I believe it can only be achieved with heavy regulation and monitoring of these tools. I think it will be a long time before our society will ever get close to having a "good AI". I believe it will be difficult for regulatory bodies to enforce the policies and regulations needed to control AI, especially with how deeply integrated big AI is already in our life.

I appreciate the sentiment, but I think that the convenience and sheer speed of AI's responses has been overlooked by the author. The uses of AI are almost limitless, and almost anything can be generated within seconds. While I would hope that most people would choose human generated over AI slop, I do not think this is the case. Not only is it difficult for many people to differentiate between human and AI content, but I would also argue that most people do not know (or care) what is going on behind the scenes and just accept AI models as wonderful tools. We cannot ignore the fact that most people using these tools may not be fully aware of the implications, but are just perfectly happy with the outputs they receive and will continue to use it.